ML, PL, QL in Markov Chain Models

In many spatial and spatial-temporal models, and more generally in models with complex dependencies, it may be too difficult to carry out full maximum-likelihood (ML) analysis. Remedies include the use of pseudo-likelihood (PL) and quasi-likelihood (QL) (also called the composite likelihood). The present paper studies the ML, PL and QL methods for general Markov chain models, partly motivated by the desire to understand the precise behaviour of the PL and QL methods in settings where this can be analysed. We present limiting normality results and compare performances in different settings. For Markov chain models, the PL and QL methods can be seen as maximum penalized likelihood methods. We find that QL is typically preferable to PL, and that it loses very little to ML, while sometimes earning in model robustness. It has also appeal and potential as a modelling tool. Our methods are illustrated for consonant-vowel transitions in poetry and for analysis of DNA sequence evolution-type models. Copyright (c) 2007 Board of the Foundation of the Scandinavian Journal of Statistics..

[1]  Erik T. Parner,et al.  A Composite Likelihood Approach to Multivariate Survival Data , 2001 .

[2]  P. Donnelly,et al.  Approximate likelihood methods for estimating local recombination rates , 2002 .

[3]  A. Hobolth,et al.  Statistical Applications in Genetics and Molecular Biology Statistical Inference in Evolutionary Models of DNA Sequences via the EM Algorithm , 2011 .

[4]  T. W. Anderson,et al.  Statistical Inference about Markov Chains , 1957 .

[5]  P. Billingsley,et al.  Statistical Methods in Markov Chains , 1961 .

[6]  A. R. de Leon.  Pairwise likelihood approach to grouped continuous model and its extension , 2005 .

[7]  M. Kimura.  Estimation of evolutionary distances between homologous nucleotide sequences. , 1981, Proceedings of the National Academy of Sciences of the United States of America.

[8]  J. Besag.  Statistical Analysis of Non-Lattice Data , 1975 .

[9]  D. Cox,et al.  A note on pseudolikelihood constructed from marginal densities , 2004 .

[10]  J. Hartigan,et al.  Asynchronous distance between homologous DNA sequences. , 1987, Biometrics.

[11]  R. Henderson,et al.  A serially correlated gamma frailty model for longitudinal count data , 2003 .

[12]  M. Kimura.  A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences , 1980, Journal of Molecular Evolution.

[13]  Samuel Karlin,et al.  A First Course on Stochastic Processes , 1968 .

[14]  Geert Molenberghs,et al.  A pairwise likelihood approach to estimation in multilevel probit models , 2004, Comput. Stat. Data Anal..

[15]  C. Bhat,et al.  A Flexible Spatially Dependent Discrete Choice Model: Formulation and Application to Teenagers’ Weekday Recreational Activity Participation , 2010 .

[16]  Øivind Skare,et al.  Pairwise likelihood inference in spatial generalized linear mixed models , 2005, Comput. Stat. Data Anal..

[17]  David J. Nott,et al.  Pairwise likelihood methods for inference in image models , 1999 .

[18]  S. Lele,et al.  A Composite Likelihood Approach to Binary Spatial Data , 1998 .

[19]  B. Kedem,et al.  Regression Theory for Categorical Time Series , 2003 .

[20]  H. Omre,et al.  Topics in spatial statistics. Discussion and comments , 1994 .

[21]  Patrick Billingsley,et al.  Statistical inference for Markov processes , 1961 .

[22]  B. Blaisdell.  A method of estimating from two aligned present-day DNA sequences their ancestral composition and subsequent rates of substitution, possibly different in the two lineages, corrected for multiple and parallel substitutions at the same site , 2005, Journal of Molecular Evolution.

[23]  R. W. Wedderburn.  Quasi-likelihood functions, generalized linear models, and the Gauss-Newton method , 1974 .